No battery Vsys issue is USB overcurrent?

Werner Almesberger werner at
Mon Jun 30 17:39:58 CEST 2008

Andy Green wrote:
> I really couldn't make a story for it, GSM power is controlled by
> another FET switch U1705 that is enabled by a PMU GPIO, there is not
> even a path to VBAT (GSM power rail) during all this.

That's not the VB->U1705->VBAT->GSM path, but the
VB->VB_GSM->U201.VBATT path. Has a nice 22uF bypass on it (C208).

This is eerily reminiscent of a problem we once had on version 1 of
the debug board: in some fraction of the boards, the EEPROM couldn't
be programmed. This turned out to be caused by a power supply drop
that occurred just around the time the FT2232 read its initial
configuration from the EEPROM, so the EEPROM was in fact programmed,
but the chip never saw the data and reverted to hardwired defaults.
The power drop was caused by excessive bypassing of the USB
downstream ports (more bypassing is better, right ? :-), which
emptied the upstream capacitor.

Yes, such things do happen in real life ;-)

> Hey I have something you'll like though Joerg, when I stick my scope
> probe on the Vsys side of the added Vsys cap, the ticking / motorboating
> noise coming from the Mic increases in amplitude :-O

All roads lead to Rome ;-)

- Werner

More information about the openmoko-kernel mailing list